Electronic Theses and Dissertation

Peningkatan konsentrasi karbon dioksida (CO₂) sebagai gas rumah kaca berkontribusi terhadap perubahan iklim sehingga diperlukan pemantauan dan peramalan emisi. Penelitian ini bertujuan menganalisis sebaran spasial, tingkat emisi, dan meramalkan konsentrasi emisi CO₂ di Kota Banda Aceh menggunakan model Generalized Space Time Autoregressive (GSTAR). Data yang digunakan berupa citra satelit Sentinel-5P periode Januari 2022–Desember 2025 yang diolah melalui Google Earth Engine (GEE) dan divisualisasikan menggunakan ArcGis. Pemodelan dilakukan pada empat kecamatan, yaitu Jaya Baru, Kuta Alam, Baiturrahman, dan Ulee Kareng, dengan bobot spasial Inverse Distance Weighting (IDW). Penentuan orde model menggunakan Minimum Information Criterion (MINIC) berdasarkan nilai Corrected Akaike Information Criterion (AICC), sedangkan evaluasi model menggunakan MAPE, MSE, dan RMSE. Hasil penelitian menunjukkan bahwa sebaran emisi CO₂ di Kota Banda Aceh didominasi Zona 2, sedangkan konsentrasi emisi selama periode 2022–2025 berada pada kisaran 500–800 ppm dengan pola fluktuatif. Model terbaik yang diperoleh adalah GSTAR(1,1) dengan nilai MAPE 10,92%–11,69%. Hasil peramalan menunjukkan bahwa konsentrasi emisi CO₂ pada periode 2026–2027 diperkirakan
tetap mengikuti pola historis, sehingga model GSTAR(1,1) layak digunakan untuk menggambarkan kecenderungan perubahan emisi CO₂ di Kota Banda Aceh.
Kata kunci: emisi CO₂, Google Earth Engine, GSTAR, Inverse Distance Weighting, peramalan spasial.

The increasing concentration of carbon dioxide (CO₂) as a greenhouse gas contributes significantly to climate change, making emission monitoring and forecasting essential. This study aimed to analyze the spatial distribution, identify the emission levels, and forecast CO₂ concentrations in Banda Aceh City using the Generalized Space Time Autoregressive (GSTAR) model. The data used were Sentinel-5P satellite imagery from January 2022 to December 2025, processed using Google Earth Engine (GEE) and visualized with ArcGis. The GSTAR model was developed using data from four districts, namely Jaya Baru, Kuta Alam, Baiturrahman, and Ulee Kareng, with the Inverse Distance Weighting (IDW) spatial weighting matrix. The model order was determined using the Minimum Information Criterion (MINIC) based on the Corrected Akaike Information Criterion (AICC), while model performance was evaluated using MAPE, MSE, and RMSE. The results showed that the spatial distribution of CO₂ emissions in Banda Aceh was dominated by Zone 2, while CO₂ concentrations during 2022–2025 ranged from 500 to 800 ppm with a fluctuating pattern. The best forecasting model was GSTAR(1,1), which achieved MAPE values ranging from 10.92% to 11.69%. The forecasting results indicated that CO₂ concentrations during 2026–2027 are expected to follow historical fluctuation patterns, suggesting that the GSTAR(1,1) model is suitable for describing future trends in CO₂ emissions in Banda Aceh City. Keywords: CO₂ emissions, GSTAR, Google Earth Engine, Inverse Distance Weighting, spatial forecasting.